Generational Extension and the Selective Turnover Coefficient Across Historical Epochs

The conventional narrative holds that modern medicine has "relaxed" natural selection by eliminating many causes of premature death. This framing, while not incorrect, misses a more fundamental demographic effect: even if selection pressures remained constant, the increase in life expectancy and shift of mortality from reproductive to post-reproductive ages would dramatically slow the rate at which selection can change allele frequencies. We derive the Selective Turnover Coefficient (d), which quantifies the effective fraction of the gene pool replaced per generation, and calculate d across historical epochs from the Paleolithic to the present. Using Coale-Demeny model life tables calibrated to paleodemographic estimates and historical vital statistics, we show that d has declined approximately 35-fold from the Neolithic (d ≈ 0.53) to 2020 (d ≈ 0.015), and approximately 44-fold from Paleolithic hunter-gatherer conditions (d ≈ 0.66). This "generational extension"—the increasing overlap between generations as lifespans lengthen—means that the timescale for allele frequency change under selection has decoupled from historically-relevant timescales. A beneficial allele with selection coefficient s = 0.01 that would reach fixation in approximately 18,000 years under Neolithic demography would require over 600,000 years under contemporary conditions.